Sunday, June 8, 2014

Brief Ontario update

A quick update incorporating the data in polls released since Friday evening, which I will analyse in greater detail on Monday morning.

The Liberals and PCs are now tied in the popular vote projection, with 37.3% apiece (or between 36% and 41%). This represents a 1.6-point drop for the Liberals and a 2.3-point gain for the Tories. The New Democrats are up 1.2 points to 19.6% (or between 18% and 21%), while the Greens were down to 4.7% (or between 3% and 6%).

In terms of seats, the Liberals dropped three to 50 (or between 43 and 59) while the PCs were unchanged at 40 (or between 33 and 48). However, the PCs' upper band increased and it now overlaps with the Liberals. The NDP was up three seats to 17, or between 11 and 20 seats.

The projection was updated with three polls, by Ipsos Reid for CTV/CP24, EKOS Research for iPolitics, and Forum Research for the Toronto Star. A series of riding polls conducted by Forum were also added to the model.

54 comments:

  1. Time to accept that the polls and analysis aren't going to call this election !!

    ReplyDelete
    Replies
    1. Alright then, Peter. We'll see you after the vote!

      Delete
    2. Nope I'll still be here because as we all know in these last few days significant swings can happen. Let's see if they do ??

      Delete
  2. What's happening in Dufferen-Caledon riding?
    Previous polls showed pcpo leading by 1.5%~3.4%...
    Then they debated again and olp(22%?) leader visited and announced GoExpress bus...and olp's sins were alleged as contravention...

    ReplyDelete
    Replies
    1. The projection is based on regional polls, not polls specific to Dufferin-Caledon.

      Delete
  3. This comment has been removed by the author.

    ReplyDelete
  4. Can you explain how you incorporate the riding level polls? According to Forums riding poll of Trinity-Spadina, the Lib-NDP spread is 1%. According to your riding projections, it is 6.5%.

    ReplyDelete
    Replies
    1. I see that is not mentioned in my methodology explanation. I will fix that.

      In short, the poll takes up part of the riding projection, but is adjusted according to how the polls have moved since the date of the riding poll. So, for example, in Trinity-Spadina the poll was done on June 1. The results of the poll are then adjusted by how the projection has moved in the region since June 1.

      Delete
    2. I also have a question regarding riding level polls.

      Do you believe Forum's riding level polls resulted in the seat projection being skewed a bit? I am wondering since Forum has only published data in 18 ridings, which means we may not know local movement in neighbouring ridings.

      For example, I see that your model now projects Burlington as 57% chance of going Liberal, while there is a 63% chance that the neighbouring Halton riding (which includes parts of Oakville and Burlington) goes PC. Forum did not poll Halton. Perhaps the Liberals have become more competitive in the region?

      Delete
  5. With the Ipsos, I get:

    43 OLP
    43 PC
    21 NDP

    With their likely-voters, I get:

    48 PC
    38 OLP
    21 NDP

    With the EKOS numbers, I get:

    44 OLP
    43 PC
    20 NDP

    With their likely-voters, I get:

    58 OLP
    36 PC
    13 NDP

    With the Forum numbers, I get:

    51 OLP
    41 PC
    15 NDP

    And finally, with the 308-aggregate, I get:

    47 OLP
    44 PC
    16 NDP

    ReplyDelete
  6. Fun election. Early polls had me at 2% (Green Party Thunder Bay-Atikokan) but now have me at 9%. I was told I did well at the 2 debates I was allowed to speak at (2 others were limited to the big 3) plus I've been aggressive with sending messages to the TV and newspaper regularly and finding myself getting a fair amount of coverage. Seeing the NDP in freefall isn't a shock as their candidate did very poorly at the debates despite nearly winning last time. The PC's shouldn't have a hope but their candidate has strong business connections (ex Chamber of Commerce president) thus the local factor is having an effect.

    Thanks for listing the riding polls Eric as I hadn't heard of that poll before. Really gives me a charge going into the final few days. We'll see if the two ads we have going this week (front of sports section, full colour) help push us up to the NDP level - if I could beat them locally it'd be like a bomb going off politically here.

    I just wish someone would run a public poll in Guelph or Dufferin-Caledon as the GPO is very strong in both and has a shot at 2 seats this time. Now _that_ would be a political bomb going off province wide and could (in this tight a race) give us the balance of power potentially.

    ReplyDelete
    Replies
    1. Best of Luck John. Hope you get in

      Delete
    2. The Greens are definitely doing better in this campaign that the last. I would not be surprised if they double their popular vote total from the last election.

      Apathy for the three major parties and their "bold" position of the school board question should make the Green Party more noticed.

      Mike Schreiner seems much more comfortable and confident in his role than he did in 2011. I could see him coming close or even causing an upset in Guelph.

      Delete
    3. John, I'm the CFO for the Green candidate down here in Windsor-Tecumseh. Wishing you luck! Let's ride that Green Wave :)

      Delete
  7. As per Peter, John whom I find myself in agreement on many more things provincially.Be great to see you beat the NDP. In my riding it is between PC and NDP. My vote has already gone to the NDP. Anything to keep Hudak from the door.

    ReplyDelete
  8. I still don't understand why the Ipsos likely voter model favours the PCs and the the Ekos likely voter model favours the Liberals. The former makes more sense to me (older people vote more often and are more likely to vote PC, younger Liberal and NDP supporters are less likely to show up).

    ReplyDelete
  9. This is absolutely fascinating. There exists the possibility that the PC party could win the popular vote with the Liberals winning slightly more seats! The Liberals could actual pull off a minority government without winning the popular vote due to the majority of their seats being won in the GTA and Toronto. This election will be won on who is able to motivate their base to vote!

    How exciting!

    ReplyDelete
    Replies
    1. That's just what happened in 1985. David Peterson's Liberals received the most votes, but got 48 seats to 52 for Frank Miller's Tories. The NDP decided it was time to change governments after 43 years of Tory rule, and backed the Grits through the accord Bob Rae signed with Peterson.

      Who knows what Horwath might do in a similar situation this time. Will she go with the party that wins the most votes or the most seats? Will she negotiate an "accord", or operate in a more traditional manner for a third party in a minority government? Or might she seek a coalition like they have in the UK currently?

      Next weekend should be very interesting.

      Delete
    2. Like 1985. If that happens, who wants to bet that Tim Hudak and Andrea Horwath will start talking?

      Delete
    3. Stevos,

      It is important to remember Miller was defeated on a confidence motion then, Peterson became premier with the support of the NDP after cobbling together an accord.

      Delete
    4. Yes, I agree, that's quite correct.

      The timing of events was interesting. Rae and Peterson negotiated the accord after the election but before the legislature reconvened and the vote on the throne speech was taken. Rae also had perfunctory discussions with the Conservatives, but it was clear early on that he was going to go with Peterson.

      Once the accord was signed, Peterson went to see the Lt-Governor and said, "Look, if the government is defeated on its throne speech and Premier Miller seeks a dissolution and a fresh election, you should know what there is an alternative, stable government that can be formed instead of having immediate elections, based on this accord."

      Peterson and Rae then combined to defeat the government-- on a vote on an opposition amendment to the throne speech declaring non-confidence in Miller's government if I recall correctly-- and the Lt-Governor swore in Peterson as Premier.

      Also if I recall correctly, Premier Miller spoke rather ungraciously in the legislature during the debate on that opposition motion. That was an unfortunate ending to his successful career in government.

      Delete
    5. It is Westminster tradition that the party with the largest number of seats is given first shot at forming government. However, it is also Westminster tradition that should that attempt fail, the second-largest party is then given a shot at forming government, rather than immediately returning to the polls.

      Delete
    6. Yes, but the tradition is also affected by the length of time the governing party has been in office post-election. If they are defeated on the Throne Speech or within weeks or a couple of months after the last election, then the second party is given a shot at forming a stable government. If the government elected governs for some time before falling, then a request to have a new election is more likely to be granted.

      Delete
    7. Stevos,

      I would be very surprised if Peterson went to Aird after the Accord was signed but, before he became premier unless Aird called upon him to do so. To offer advice to the Crown before becoming a minister is inappropriate and unconstitutional unless the advice is requested. Aird as a well respected lawyer would have understood this.

      Paul,
      The idea that the second party has a "right" to form government if the largest party is unable to gain or hold confidence is very flimsy. It happened during King-Byng, Ontario 1985 and Australia in 1975 but, only on the condition that a double dissolution be requested. Ontario in 1985 is the only successful use of the second party gets to form government theory. The Lascelles principles discuss when the Sovereign may deny a request for dissolution, in so doing they help outline the minimum needed or expected of a government. 1. The Parliament is still viable and capable of doing its job. 2. A general election would be detrimental to the national economy (some believe this clause is no longer in effect) 3. he (the Sovereign or Governor) could rely on finding another Prime Minister who could carry on his Government, for a reasonable period, with a working majority in the House of Commons.

      The Crown chooses the government and at the end of the day The Monarch or Governor has sole discretion whom they wish to appoint as premier.

      If on Friday we are left with a hung Parliament with Horwath refusing to support either the Liberals or Tories on supply the likely outcome would be a dissolution once the budget was defeated. If however, Parliament is still viable and able to pass supply through some combination of MPPs the Government may last for some time. The idea that time has a determinant effect upon whether the leader of the opposition is asked to become premier is I think possible but, not necessarily borne out by historical examples. The Crown cares more about stability and supply then how long ago the last general election took place. Certainly in 1985 time may have been a factor and Peterson was appointed premier as was the case in 1926. However, it is important to remember Malcolm Fraser also became PM after Gough Whitlam was dismissed 18 months after the last election and an election called not because enough time had elapsed but, in order to break the parliamentary deadlock that had precipitated the crisis.

      Delete
  10. Ipsos Reid has been polling on their Opt-in panel (a voluntary group) so the sampling may not be as "random".

    ReplyDelete
    Replies
    1. as random as what?

      IVR is just another name for Robo call...

      Do you or anyone you know respond to Robo-calls? or even answers them?

      Anyone who goes through a complete IVR call is already a specific slice of the population that is not at all random as far as the Voting Universe.
      is?

      Delete
    2. The pollsters are aware of this and attempt to factor in its effects when they computer their poll results. Example: if I believe that the supporters of party X are going to respond to half of my calls, but that they only represent 30% of the actual electorate, then if half of the people I call say they support party X, I will record party X at 30%, not 50%. If only 25% of the people I call say they support party X, I will similarly adjust, and ave party X at 15%. The accuracy of these models is a significant factor in the accuracy of the pollster.

      Delete
  11. The Globe and Mail and Toronto Sun have endorsed the PCs to form Ontario's next government!

    The Toronto Star (unsurprisingly) is endorsing Wynne to be re-elected.

    ReplyDelete
    Replies
    1. Bluntly Formerly the ONLY endorsement that matters is the electors. Stop grasping straws !

      Delete
    2. When Mr. Hudak wins a majority on Thursday how will you eat your words Peter; newsprint or bible paper?

      Hahaha, just kidding of course.

      I do think momentum has shifted slightly towards the Tories over the last few days; 67% of those polled think it time for a change at Queen's Park, small bump in the polls for the Tories, Media endorsements favouring the PCs 2:1 over the Liberals thus far, more bad news for the Liberals regarding gas plants today.

      Delete
    3. Well, during the last election (Oct 2011) nearly two-thirds of the electorate thought it was time for a change at Queen's Park. Then they looked at their options and the Liberals were re-elected.

      If 33% of the electorate are okay with the status quo. It takes maybe another 3-4% of the electorate who may hold their nose and vote Liberal once again.

      Delete
    4. Not surprising where the Sun places its endorsement. AS for the Globe, their support for Hudak is very lukewarm and qualified. They actually say they want a minority government with him as Premier. How does one vote for that? Conspire with the neighbours to sprinkle votes among the parties up and down the block ? They're advocating an option that doesn't exist.

      Delete
    5. And the real problem is ?

      Trend in voter turnout

      1990: 64.4% of eligible voters
      1995: 62.9%
      1999: 58.3%
      2003: 56.8%
      2007: 52.1%
      2011: 48.2%

      Source: Elections Ontario

      Delete
    6. "67% of those polled think it time for a change at Queen's Park"

      You can put me in that 67%. However, that simply means I'm not voting Liberal. It doesn't mean I'm voting Conservative. By rights, I should be. Perhaps the PC party can do some soul searching and think about why so many of us who have voted PC in the past are unwilling to do so in this election.

      Delete
    7. So now we have less than half of the voters directing to course of our democracy/province ?? Somehow that does not fill me with confidence !!

      Delete
    8. It certainly should not imbue you with confidence Peter. Disgraceful to say the least that turnout is so low. A number of factors are probably to blame such as; time, poor civics education (especially for immigrants), inconvenience, politicians and political parties themselves etc...

      I question whether Elections Ontario's or Elections BC or Elections Canada estimates of eligible voters is overstated. Many immigrants to Canada do not take out Canadian citizenship for a host of reasons. I wonder if low turnout is partially the result of a discrepancy between population and eligible voters.

      Delete
    9. Actually, Stevos, conspiring with the neighbours won't help ensure a minority government -- you'd have to conspire with enough people in other ridings to guarantee a sufficient number of ridings going to each party to deny anyone a majority. If the conspiracy stays within a riding, you can arrange to give the new MP a small margin of victory, but that won't change the overall seat calculations.

      Delete
    10. I don't think there is any answer to it bede short of compulsory voting. The trend numbers over the years are quite clear. Fewer and fewer give a shit !!

      Delete
    11. Pardon me Peter but is your answer to force all those who "don't give a shit" to vote?

      Do you think forcing low-information voters to the polls will produce a more effective government?

      Colour me skeptical. Those who CHOOSE not to vote are in fact making a choice. I respect that choice.

      To all those who say you can't complain if you don't vote I will present the counter-argument.

      If you pay taxes than you have every right to complain about those dollars being wasted, regardless of who you voted for (or not at all).

      Many might say that wars were fought to give us the right to vote. The truth of this statement is debatable, but if that is so than they also fought so that you have the right NOT to vote.

      The idea of forcing people who couldn't otherwise be bothered to vote is not an answer to a more representative government. It is a recipe for worse outcomes, as well as an unneeded intrusion by the state into the lives of it's citizens.

      Delete
    12. "Pardon me Peter but is your answer to force all those who "don't give a shit" to vote?"

      Absolutely NOT !! Political parties actually need to change their approach and personna to try and engage these people who don't care about voting. Also look at things like mobile polls or internet voting. Sure they all have problems but the slumping participation numbers are scary !!

      Delete
    13. Citizenship is about more than paying your taxes. It is about contributing to your country and society.

      Just because a country or province have compulsory voting does not force anyone to vote; one may decline the ballot or spoil the ballot both have the same effect as not voting but, also have the effect of requiring participation in some form.

      I do not think there is any evidence to suggest compulsory voting in Canada or Ontario would produce worse outcomes. Australia where voting is mandatory appears to muddle along as well as Canada if not slightly better. Also it is important to remember "compulsory voting" is not coercion-nobody will stick a gun to your back and walk you to the polling booth. The penalty for not voting is a $50 fine in Australia.

      Finally, I really do not think you can say people died to give anyone the right not to vote. People died in various wars to give the right of free speech, democracy et cetera but, they did not fight simply for you or I to be lazy. In order to equate democracy or not voting with a "right" the act of not voting must be consciously thought out. If one chooses not to vote to make a statement fine but, if one does not vote due to laziness that is not what people fought and died for.

      Delete
  12. Also about the supposed decline of the NDP: Many users on http://www.electionprediction.org/ are suggesting that the NDP will win Kingston and the Islands. This would be rather surprising given that the Liberals under John Gerretsen have won this riding many times.

    ReplyDelete
    Replies
    1. I wouldn't read too much into the comments on electionprediction.org regarding Kingston-and-the-Islands. Many appear to be written by NDP partisans.

      For example: "Wheels have come off the Liberal campaign - NDP campaign is firing on all cylinders". But, of course the author does not present any evidence even of the qualitative variety to back his statement.

      Delete
    2. As someone that spends a great deal of time in Kingston for school, I don't think the NDP can pull it off. Second place is likely in Kingston and the Islands, but it's more likely that Sophie Kiwala will win the riding.

      Delete
    3. I'd say the NDP would have a chance in Kingston and the Islands if they are polling in the high 20s like they were a few months ago. A center-left urban seat without an incumbent is a perfect place for Horwath's NDP to gain seats.

      However, with the NDP's current poll numbers, I doubt they would win this riding. Eastern Ontario is currently their weakest region too.

      Delete
  13. The media endorsements section on Wikipedia is interesting - most papers are endorsing the PCs this time, rather than the Liberals like last time. Has there ever been an election in which the majority of newspapers did NOT endorse the party that eventually won?

    I'm looking at other Canadian election articles on Wikipedia to try to find the answer, but only some of them have a newspaper endorsements section.

    ReplyDelete
    Replies
    1. I'm guessing the 1990 election would be an example of an election where the majority of newspapers doesn't endorse the winning party.

      At this moment, it seems like only the Globe and Mail and Windsor Star shifted their support from OLP to PC.

      On another note, I felt the Globe and Mail's endorsement of a Hudak is a cop out. (Similarly, they endorsed a Chretien minority back in 1993).

      How does one vote for a Hudak minority? Wouldn't a Hudak minority mean a legislative stalemate? The Globe is critical of the NDP, yet they would want them to continue calling the shots for years to come?

      Delete
    2. The Globe & Mail had an interesting feature on their website where they listed their previous endorsements since the 80s. They endorsed the Liberals in 81, 85 and 90 who didn't win. But then they endorsed the winning candidate each time after that, except 2007 where they endorsed both McGuinty *and* Tory, so I guess they get a half point on that.

      Delete
    3. The KW Record endorsed none and stated it is the only time they know of where that has been the case. The Globe enforcement is the weakest enforcement I've ever seen in my life. Neither the Star nor the National Post endorsements are surprising (though the Star is perhaps slightly surprising in that they could credibly have endorsed the NDP, had Horvath done a better job in her campaign (I have never seen a weaker closing statement than the one by Horvath in the debate: she didn't have a clear statement; she was making it up as she went with plenty of ums and errs. Wow. I couldn't believe it!)).

      Delete
    4. I think Paul that what we are seeing is three weak parties squabbling over the results of poor campaigns. Doesn't inspire confidence ! I wish in a way that the KW Record approach would apply to all papers. Their job is, after all, to report the news not create the news !!

      Delete
    5. The weak endorsements are showing why I'm seeing some Green growth here and why so many (even some PC's and Liberals) are hoping for a Green breakthough this time (they want 1 or 2 seats at most of course).

      We really need change to the system as a whole though. So many feel disengaged. We need fewer signs and more discussion. Less money and more thought. Sadly, the current method has worked well for the big 2 and corporations thus making it very, very hard to change anything.

      A good first step would be to do 'instant runoffs' so 50%+1 is needed to win a seat.

      Delete
  14. The Globe had a link on their website listing their past endorsements back to the early 1980s and they seem to have backed the winner in all prior elections except for 2.
    http://www.theglobeandmail.com/news/politics/ontario-election-globe-endorsements-from-1981-to-present/article19021451/
    As for the other papers, no idea.
    I wonder how influential these endorsements are though. I would imagine less so than in the past given the rise of the internet (for those who want to read multiple papers), the decline in newspaper readership.

    ReplyDelete

COMMENT MODERATION POLICY - Please be respectful when commenting. If choosing to remain anonymous, please sign your comment with some sort of pseudonym to avoid confusion. Please do not use any derogatory terms for fellow commenters, parties, or politicians. Inflammatory and overly partisan comments will not be posted. PLEASE KEEP DISCUSSION ON TOPIC.